Remote controlled stereoscopic camera devices are widely used for advanced remote control of equipment (e.g. robots, surgical systems, etc.) and surveillance.
The stereoscopic camera device comprises two cameras, displaced horizontally from one another, which are used to obtain two different views on a scene. The images from the two cameras are shown to respective eyes of a remote operator using a display device to create an impression of depth, i.e. a 3D image. When the operator turns his/her head, the cameras should also rotate to observe the scene from a different direction.
The time between detecting movement of a head of the operator and showing scene from right direction is called latency. The latency time has three main contributions: image processing delay, motor reaction time and network latency.
High latency destroys immersion effect for the operator, whereby many stereoscopic systems have a strong emphasis on achieving low latency. Image processing delay can be reduced by using light weight encoding algorithms or by increasing computational power of encoding hardware. Stronger motors are used to combat motor reaction time. Network latency is kept low by keeping operator relatively close to the cameras.